Audio Video Transport WG Q. Xie Internet-Draft D. Pearce Expires: June 11, 2003 Motorola December 11, 2002 RTP Payload Format for ETSI ES 202 050 Distributed Speech Recognition Encoding draft-xie-avt-dsr-es202050-00.txt Status of this Memo This document is an Internet-Draft and is in full conformance with all provisions of Section 10 of RFC2026. Internet-Drafts are working documents of the Internet Engineering Task Force (IETF), its areas, and its working groups. Note that other groups may also distribute working documents as Internet- Drafts. Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress." The list of current Internet-Drafts can be accessed at http:// www.ietf.org/ietf/1id-abstracts.txt. The list of Internet-Draft Shadow Directories can be accessed at http://www.ietf.org/shadow.html. This Internet-Draft will expire on June 11, 2003. Copyright Notice Copyright (C) The Internet Society (2002). All Rights Reserved. Abstract This document specifies an RTP payload format for encapsulating ETSI Standard ES 202 050 advanced front-end signal processing feature streams for distributed speech recognition (DSR) systems. Xie & Pearce Expires June 11, 2003 [Page 1] Internet-Draft RTP Payload Format for ES202050 DSR December 2002 Table of Contents 1. Conventions . . . . . . . . . . . . . . . . . . . . . . . . . 3 2. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3 2.1 ETSI ES 202 050 DSR Front-end Codec . . . . . . . . . . . . . 3 3. ES 202 050 DSR RTP Payload Format . . . . . . . . . . . . . . 4 3.1 Consideration on Number of FPs in Each RTP Packet . . . . . . 5 3.2 Support for Discontinuous Transmission . . . . . . . . . . . . 5 4. Frame Pair Formats . . . . . . . . . . . . . . . . . . . . . . 5 4.1 Format of Speech and Non-speech FPs . . . . . . . . . . . . . 5 4.2 Format of Null FP . . . . . . . . . . . . . . . . . . . . . . 7 4.3 RTP header usage . . . . . . . . . . . . . . . . . . . . . . . 7 5. DSR MIME Type Registration . . . . . . . . . . . . . . . . . . 8 5.1 Mapping MIME Parameters into SDP . . . . . . . . . . . . . . . 9 6. Security Considerations . . . . . . . . . . . . . . . . . . . 9 7. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 10 Normative References . . . . . . . . . . . . . . . . . . . . . 10 Informative References . . . . . . . . . . . . . . . . . . . . 10 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . 11 Full Copyright Statement . . . . . . . . . . . . . . . . . . . 12 Xie & Pearce Expires June 11, 2003 [Page 2] Internet-Draft RTP Payload Format for ES202050 DSR December 2002 1. Conventions The keywords MUST, MUST NOT, REQUIRED, SHALL, SHALL NOT, SHOULD, SHOULD NOT, RECOMMENDED, NOT RECOMMENDED, MAY, and OPTIONAL, when they appear in this document, are to be interpreted as described in [4]. The following acronyms are used in this document: DSR - Distributed Speech Recognition ETSI - the European Telecommunications Standards Institute FP - Frame Pair DTX - Discontinuous Transmission 2. Introduction Distributed speech recognition (DSR) technology is intended for a remote device acting as a thin client, also known as the front-end, to communicate with a speech recognition server, also called a speech engine, over a network connection to obtain speech recognition services. More details on DSR over Internet can be found in [7]. To achieve interoperability with different client devices and speech engines, the first ETSI standard DSR front-end ES 201 108 was published in early 2000 [8], and an RTP packetization for ES 210 108 frames is defined in [7] in IETF. In ES 202 050 [1], ETSI issues another standard for an Advanced DSR front-end that is meant to provide substantially improved recognition performance in background noise. The codecs in ES 202 050 uses a different frame format from that of ES 201 108 and the two do not inter-operate with each other. Thus, this document defines a separate RTP packetization for ES 202 050 front-end. 2.1 ETSI ES 202 050 DSR Front-end Codec Some relevant characteristics of ES 202 050 DSR front-end codec are summarized below. The coding algorithm, a standard mel-cepstral technique common to many speech recognition systems, supports three raw sampling rates: 8 kHz, 11 kHz, and 16 kHz. The mel-cepstral calculation is a frame- based scheme that produces an output vector every 10 ms. After calculation of the mel-cepstral representation, the representation is first quantized via split-vector quantization to reduce the data rate of the encoded stream. Then, the quantized Xie & Pearce Expires June 11, 2003 [Page 3] Internet-Draft RTP Payload Format for ES202050 DSR December 2002 vectors from two consecutive frames are put into an FP, as described in more detail in Section 4.1. 3. ES 202 050 DSR RTP Payload Format An ES 202 050 DSR RTP payload datagram consists of a standard RTP header [2] followed by a DSR payload. The DSR payload itself is formed by concatenating a series of ES 202 050 DSR FPs (defined in Section 4). FPs are always packed bit-contiguously into the payload octets beginning with the most significant bit. For ES 202 050 front-end, the size of each FP is 96 bits or 12 octets (see Sections 4.1 and 4.2). This ensures that a DSR payload will always end on an octet boundary. The following example shows a DSR RTP datagram carrying a DSR payload containing three 96-bit-long FPs (bit 0 is the MSB): 0 1 2 3 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ \ \ / RTP header in [RFC1889] / \ \ +=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+ | | + + | FP #1 (96 bits) | + + | | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | | + + | FP #2 (96 bits) | + + | | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | | + + | FP #3 (96 bits) | + + | | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ Xie & Pearce Expires June 11, 2003 [Page 4] Internet-Draft RTP Payload Format for ES202050 DSR December 2002 3.1 Consideration on Number of FPs in Each RTP Packet The number of FPs per payload packet should be determined by the latency and bandwidth requirements of the DSR application using this payload format. In particular, using a smaller number of FPs per payload packet in a session will result in lowered bandwidth efficiency due to the RTP/UDP/IP header overhead, while using a larger number of FPs per packet will cause longer end-to-end delay and hence increased recognition latency. Furthermore, carrying a larger number of FPs per packet will increase the possibility of catastrophic packet loss; the loss of a large number of consecutive FPs is a situation most speech recognizers have difficulty dealing with. It is therefore RECOMMENDED that the number of FPs per DSR payload packet be minimized, subject to meeting the application's requirements on network bandwidth efficiency. RTP header compression techniques, such as those defined in [9] and [10], should be considered to improve network bandwidth efficiency. 3.2 Support for Discontinuous Transmission The DSR RTP payloads may be used to support discontinuous transmission (DTX) of speech, which allows that DSR FPs are sent only when speech has been detected at the terminal equipment. In DTX a set of DSR frames coding an unbroken speech segment transmitted from the terminal to the server is called a transmission segment. A DSR frame inside such a transmission segment can be either a speech frame or a non-speech frame, depending on the nature of the section of the speech signal it represents. The end of a transmission segment is determined at the sending end equipment when the number of consecutive non-speech frames exceeds a pre-set threshold, called the hangover time. A typical value used for the hangover time is 1.5 seconds. After all FPs in a transmission segment are sent, the front-end SHOULD indicate the end of the current transmission segment by sending one or more Null FPs (defined in Section 4.2). 4. Frame Pair Formats 4.1 Format of Speech and Non-speech FPs Similar to the frame pairing format defined in Section 7.2.4 in [1], pairs of the quantized 10ms mel-cepstral front-end frames, of 44 bits each, MUST be grouped together and protected with a 4-bit CRC. Xie & Pearce Expires June 11, 2003 [Page 5] Internet-Draft RTP Payload Format for ES202050 DSR December 2002 Together, these two front-end frames and the CRC field form a 92-bit long Frame-pair (FP): 0 1 2 3 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Frame #1 (44 bits) | + +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | | Frame #2 (44 bits) | +-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+ | | CRC |0|0|0|0| +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ Therefore, each FP represents 20ms of original speech. Note, as shown above, each FP MUST be padded with 4 zeros to the LSB 4 bits of the last octet in order to make the FP aligned to the 32-bit word boundary. This makes the total size of an FP 96 bits, or 12 octets. Note, this padding is separate from padding indicated by the P bit in the RTP header. The 4-bit CRC MUST be calculated using the formula defined in 7.2.4 in [1]. To be consistent with the bit-order used in [1], the following mel- cepstral frame formats MUST be used when forming an FP: Frame #1 in FP: =============== (MSB) (LSB) 0 1 2 3 4 5 6 7 +-----+-----+-----+-----+-----+-----+-----+-----+ : idx(2,3) | idx(0,1) | Octet 1 +-----+-----+-----+-----+-----+-----+-----+-----+ : idx(4,5) | idx(2,3) (cont) : Octet 2 +-----+-----+-----+-----+-----+-----+-----+-----+ | idx(6,7) |idx(4,5)(cont) Octet 3 +-----+-----+-----+-----+-----+-----+-----+-----+ idx(10,11)| VAD | idx(8,9) | Octet 4 +-----+-----+-----+-----+-----+-----+-----+-----+ : idx(12,13) | idx(10,11) (cont) : Octet 5 +-----+-----+-----+-----+-----+-----+-----+-----+ | idx(12,13) (cont) : Octet 6/1 +-----+-----+-----+-----+ Xie & Pearce Expires June 11, 2003 [Page 6] Internet-Draft RTP Payload Format for ES202050 DSR December 2002 Frame #2 in FP: =============== (MSB) (LSB) 0 1 2 3 4 5 6 7 +-----+-----+-----+-----+ : idx(0,1) | Octet 6/2 +-----+-----+-----+-----+-----+-----+-----+-----+ | idx(2,3) |idx(0,1)(cont) Octet 7 +-----+-----+-----+-----+-----+-----+-----+-----+ : idx(6,7) | idx(4,5) | Octet 8 +-----+-----+-----+-----+-----+-----+-----+-----+ : idx(8,9) | idx(6,7) (cont) : Octet 9 +-----+-----+-----+-----+-----+-----+-----+-----+ | idx(10,11) | VAD |idx(8,9)(cont) Octet 10 +-----+-----+-----+-----+-----+-----+-----+-----+ | idx(12,13) | Octet 11 +-----+-----+-----+-----+-----+-----+-----+-----+ The usage of the index fields and 'VAD' flag are defined in [1] and their value is only set and examined by the codecs in the front-end client and the recognizer. 4.2 Format of Null FP A Null FP for the ES 202 050 front-end codec is defined by setting the content of the first and second frame in the FP to null (i.e., filling the first 88 bits of the FP with 0's). The 4-bit CRC MUST be calculated the same way as described in 6.2.4 in [1], and 4 zeros MUST be padded to the end of the Null FP to made it 32-bit word aligned. 4.3 RTP header usage The format of the RTP header is specified in [2]. This payload format uses the fields of the header in a manner consistent with that specification. The RTP timestamp corresponds to the sampling instant of the first sample encoded for the first FP in the packet. The timestamp clock frequency is the same as the sampling frequency, so the timestamp unit is in samples. As defined by ES 202 050 front-end codec, the duration of one FP is 20 ms, corresponding to 160, 220, or 320 encoded samples with sampling rate of 8, 11, or 16 kHz being used at the front-end, respectively. Thus, the timestamp is increased by 160, 220, or 320 for each consecutive FP, respectively. Xie & Pearce Expires June 11, 2003 [Page 7] Internet-Draft RTP Payload Format for ES202050 DSR December 2002 The DSR payload for ES 202 050 front-end codes is always an integral number of octets. If additional padding is required for some other purpose, then the P bit in the RTP in the header may be set and padding appended as specified in [2]. The RTP header marker bit (M) should be set following the general rules defined in [6]. The assignment of an RTP payload type for this new packet format is outside the scope of this document, and will not be specified here. It is expected that the RTP profile under which this payload format is being used will assign a payload type for this encoding or specify that the payload type is to be bound dynamically. 5. DSR MIME Type Registration Media Type name: audio Media subtype name: dsr-es202050 Required parameters: none Optional parameters for RTP mode: rate: Indicates the sample rate of the speech. Valid values include: 8000, 11000, and 16000. If this parameter is not present, 8000 sample rate is assumed. maxptime: The maximum amount of media which can be encapsulated in each packet, expressed as time in milliseconds. The time shall be calculated as the sum of the time the media present in the packet represents. The time SHOULD be a multiple of the frame pair size (i.e., one FP == 20ms). If this parameter is not present, maxptime is assumed to be 80ms. Note, since the performance of most speech recognizers are extremely sensitive to consecutive FP losses, if the user of the payload format expects a high packet loss ratio for the session, it MAY consider to explicitly choose a maxptime value for the session that is shorter than the default value. ptime: see RFC2327 [5]. Encoding considerations: This type is defined for transfer via RTP [2] as described in Sections 3 and 4 of RFC XXXX. Xie & Pearce Expires June 11, 2003 [Page 8] Internet-Draft RTP Payload Format for ES202050 DSR December 2002 Security considerations: See Section 6 of RFC XXXX. Person & email address to contact for further information: Qiaobing.Xie@motorola.com Intended usage: COMMON. It is expected that many VoIP applications (as well as mobile applications) will use this type. Author/Change controller: * Qiaobing.Xie@motorola.com * IETF Audio/Video transport working group 5.1 Mapping MIME Parameters into SDP The information carried in the MIME media type specification has a specific mapping to fields in the Session Description Protocol (SDP) [5], which is commonly used to describe RTP sessions. When SDP is used to specify sessions employing ES 201 018 DSR codec, the mapping is as follows: o The MIME type ("audio") goes in SDP "m=" as the media name. o The MIME subtype ("dsr-es202050") goes in SDP "a=rtpmap" as the encoding name. o The optional parameter "rate" also goes in "a=rtpmap" as clock rate. o The optional parameters "ptime" and "maxptime" go in the SDP "a=ptime" and "a=maxptime" attributes, respectively. Example of usage of ES 202 050 DSR: m=audio 49120 RTP/AVP 101 a=rtpmap:101 dsr-es202050/8000 a=maxptime:40 6. Security Considerations Implementations using the payload defined in this specification are subject to the security considerations discussed in the RTP specification [2] and the RTP profile [6]. This payload does not specify any different security services. Xie & Pearce Expires June 11, 2003 [Page 9] Internet-Draft RTP Payload Format for ES202050 DSR December 2002 7. Acknowledgments The design presented here is based on that of [7]. Normative References [1] European Telecommunications Standards Institute (ETSI) Standard ES 202 050, "Speech Processing, Transmission and Quality Aspects (STQ); Distributed Speech Recognition; Front-end Feature Extraction Algorithm; Compression Algorithms", Ver. 1.1.1 (http://pda.etsi.org/pda/home.asp?wki_id=6402), October 2002. [2] Schulzrinne, H., Casner, S., Frederick, R. and V. Jacobson, "RTP: A Transport Protocol for Real-Time Applications", RFC 1889, January 1996. [3] Bradner, S., "The Internet Standards Process -- Revision 3", BCP 9, RFC 2026, October 1996. [4] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, March 1997. [5] Handley, M. and V. Jacobson, "SDP: Session Description Protocol", RFC 2327, April 1998. [6] Schulzrinne, H. and S. Casner, "RTP Profile for Audio and Video Conferences with Minimal Control", draft-ietf-avt-profile-new- 12.txt (work in progress), November 2001. Informative References [7] Xie, Q., "RTP Payload Format for ETSI ES 201 108 Distributed Speech Recognition Encoding", draft-ietf-avt-dsr-04 (work in progress), October 2002. [8] European Telecommunications Standards Institute (ETSI) Standard ES 201 108, "Speech Processing, Transmission and Quality Aspects (STQ); Distributed Speech Recognition; Front-end Feature Extraction Algorithm; Compression Algorithms", Ver. 1.1.2, http://webapp.etsi.org/pda/home.asp?wki_id=9948, April 2000. [9] Casner, S. and V. Jacobson, "Compressing IP/UDP/RTP Headers for Low-Speed Serial Links", RFC 2508, February 1999. [10] Bormann, C., Burmeister, C., Degermark, M., Fukushima, H., Hannu, H., Jonsson, L-E., Hakenberg, R., Koren, T., Le, K., Liu, Z., Martensson, A., Miyazaki, A., Svanbro, K., Wiebke, T., Xie & Pearce Expires June 11, 2003 [Page 10] Internet-Draft RTP Payload Format for ES202050 DSR December 2002 Yoshimura, T. and H. Zheng, "RObust Header Compression (ROHC): Framework and four profiles: RTP, UDP, ESP, and uncompressed", RFC 3095, July 2001. Authors' Addresses Qiaobing Xie Motorola, Inc. 1501 W. Shure Drive, 2-F9 Arlington Heights, IL 60004 US Phone: +1-847-632-3028 EMail: qxie1@email.mot.com David Pearce Motorola Labs UK Research Laboratory Jays Close Viables Industrial Estate Basingstoke, HANTS RG22 4PD UK Phone: +44 (0)1256 484 436 EMail: bdp003@motorola.com Xie & Pearce Expires June 11, 2003 [Page 11] Internet-Draft RTP Payload Format for ES202050 DSR December 2002 Full Copyright Statement Copyright (C) The Internet Society (2002). All Rights Reserved. This document and translations of it may be copied and furnished to others, and derivative works that comment on or otherwise explain it or assist in its implementation may be prepared, copied, published and distributed, in whole or in part, without restriction of any kind, provided that the above copyright notice and this paragraph are included on all such copies and derivative works. However, this document itself may not be modified in any way, such as by removing the copyright notice or references to the Internet Society or other Internet organizations, except as needed for the purpose of developing Internet standards in which case the procedures for copyrights defined in the Internet Standards process must be followed, or as required to translate it into languages other than English. The limited permissions granted above are perpetual and will not be revoked by the Internet Society or its successors or assigns. This document and the information contained herein is provided on an "AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Acknowledgement Funding for the RFC Editor function is currently provided by the Internet Society. Xie & Pearce Expires June 11, 2003 [Page 12]